[Mrs Vampire]

Florite paints have provided all the stenographic analysis and colour matching , even visiting the Tank with their electronic equipment to ensure we had it right.
All of the paints we are testing and producing are available from them . That includes all of those on Mike Starmers chipset sheet.
where I am at with the paint colours.
picture 1 from the left KG3 (Australian) KGJ from my best sample to date Lower rectangular plate OD9 from TM9 Ordinance , Small white sample above that Inside White for Stuart spectrographed from pristine sample, happy to say 100% on that one, Upper square sample Jacques colour match of his KGJ sample the light green sample is the primer colour for the Stuart ( other than the very dark grey base primer on some other parts and areas. )
Next is the KG3 commonly used in Australia compared to Mike Starmers chipset. Mike is a very highly regarded colour expert from the UK. I note that the UK version of KG3 is considerably Browner than the Australian version. This I think lends weight to my theory that Australian KG3 is the third toning down of KGJ ( First late 1942 , then mid 1943 and again late 1943 ) But unless Mike has it completely wrong the UK and Australian KG3 are very different colours.
third photo is my latest best shot at KGJ (1942) compared with Mikes chipset. It most closely approximated the British version of OD ( but note that colour is somewhat different to American OD ) But really its like none of the UK colours.
Finally on the Desert yellow "light Stone " as seen on some 25pdrs for example the one in the AWM and the one at Bandiana. The Australian examples seem to be quite different to the one in the Imperial War Museum and on other UK based examples. I am able to get a chip of that and will have it tested and post a sample when I have it.

Mike the files have the Chief of engineering in the Army referring to Khaki Green J as does Dakin and others until the final act of despair by the army in November of 1943 where he declared KG3 would be the standard colour and all attempts at disruptive schemes would be abandoned. Dakin agreed with him. The Army had problems firstly applying the disruptive patterns as per orders, then in parking the vehicles so it was effective. If the vehicles were incorrectly placed the disruptive because of the recognizable pattern became easier to spot not harder.
After observing US vehicles painted uniformly olive drab in joint exercises with the Australians Dakin and the army were convinced it was the easiest most effective way to camouflage. The decision to adopt KG3 does not refer to what that meant in terms of actual colour. Looking at Mike Starmers chip it is clear the Australian colour is much greener than the UK version. Mike Starmers has observed the UK KG3 was not a part of the BS chipset rather a colour developed by the army over time.
Given there is little resemblance between the KG3 UK and the Australian KG3 in am inclined to the idea that this is the third version of KGJ . Though others have pointed to references to KG3 by the Australian army prior to November 1943. That , in my view is as you suppose the britisification of Australian terminology by some, and not a reflection of what was going on with the boffins who were mixing up the brew and those in the senior ranks who were getting exasperated with all the changes and near impossibility to get uniformity of application and use out in the field.
From my reading of the files I still stand by my position the only paints colours available were those approved by the military . The repair of motor cars being one of the few exceptions where a civilian could get paint for other than a defense authorized purpose. The paint supply committee held tight reign on the rationing of paint and pigment.
The issue with differences in colour seems to narrow down to paint that was made in Melbourne which faded very quickly and was considered unsuitable. It took a little while for that situation to be resolved but explains that in black and white photos there appears to be some widely different colours.